S8 Forestry and mercury: Defining the connection

Tuesday, 26 July, 2011

TS8-P1 — 11:00-12:00 and 17:30-18:30
Authors: ALLAN, Craig1, HEYES, Andrew 2, MAKERETH, Robert J.3
(1) University of North Carolina at Charlotte, cjallan@uncc.edu; (2) Chesapeake Biological Laboratory; (3) Ontario Ministry of Natural Resources.

A spatial survey of watershed water quality in eighty-one catchments varying from 26 ha to higher order rivers draining several hundred km of boreal forest on the north shore of Lake Superior was conducted from August 2008 to September 2010. The watersheds examined in this study exhibited variable amounts and variable temporal histories of logging disturbance as well as sites that drained urbanized regions in and around the city of Thunder Bay, Ontario. In addition to the extensive spatial survey of surface waters from this region, a subset of these same watersheds along with throughfall (deciduous and conifer source), soil water and shallow and deep groundwaters were also collected and analyzed monthly. Samples were analyzed for methyl Hg (MeHg) and total Hg (THg), DOC, nutrients and major ions. A spectral characterization of DOM involving excitation emission matrix fluorescence spectroscopy combined with PARAFAC analysis was carried out on subsamples from these same sites. Several significant differences in DOM quality were measured between sampling locations. However, no relationship between Hg or MeHg concentrations and catchment disturbance or other catchment characteristics (eg. wetland area) is evident from our data set. The only significant relationship found between DOM and Hg occurred between DOC and THg and more significantly Abs 254nm and THg. However, both of these relationships were weak and the form of the relationship differed significantly from other studies (eg. Dittman et al. 2009.) Our data suggests: 1) This area of the Canadian Precambrian Shield is less sensitive to mercury mobilization from forest management activities than boreal areas in Finland and Sweden; 2) The narrow range of Hg concentrations observed in streams in rivers from this region corroborate the results from intensively studied logging experiments where enhanced Hg transport was driven by increased water fluxes rather than concentration increases; and 3) Hg appears to be supply limited in relationship to carbon availability and carbon/Hg relations are less defined in comparison to other regions.

TS8-P2 — 11:00-12:00 and 17:30-18:30
Authors: OBRIST, Daniel1, JOHNSON, Dale2, POKHAREL, Ashok3, LINDBERG, Steve4, LUO, Yiqi5, HARARUK, Okesandra5
(1) Desert Research Institute, dobrist@dri.edu; (2) University of Nevada, Reno; (3) Desert Research Insitute; (4) Oak Ridge National Laboratory - Emeritus; (5) University of Oklahoma, Norman;

Results from a systematic investigation of Hg concentrations across 14 forests in the United States show highest concentrations of total Hg in litter, strongly enriched compared to aboveground tissues and indicative of substantial post-depositional sorption of Hg. Soil Hg concentrations were lower than in litter, with highest concentrations in surface soils. Aboveground tissues showed no clear spatial distribution patterns, likely due to 17 different tree species present across sites. Litter and soil Hg concentrations strongly correlated with carbon (C), latitude, precipitation, and clay, which together explained up to 94% of concentration variability. We observed strong latitudinal increases in Hg in soils and litter, in contrast to inverse latitudinal gradients of atmospheric deposition measures. Based on a multi-regression model, we created a distribution map of Hg in surface soils of the United States. Concentrations of methyl-Hg were strongly related to the presence of total Hg, and hence sites with high concentrations and pools of total Hg correspondingly showed high methyl-Hg concentrations and pools.

Soil and litter Hg concentrations were closely linked to C contents, consistent with well-known associations between organic matter and Hg, and we propose that C shapes the distribution of Hg in forests at continental scales. The consistent link between C and Hg distribution may reflect a long-term legacy whereby old, C-rich soil and litter layers sequester atmospheric Hg depositions over long time periods. Total pools of Hg were calculated for all above-ground and below-ground pools and we discuss total ecosystem pools in relation to dry and C pools.

A series of experimental field and laboratory studies were performed to assess fate of C sequestered in litter and soils pools upon C mineralization. Our results consistently show that a fraction of Hg bound to soils and litter is subject to gaseous evasion upon C mineralization. In litter, we found that gaseous losses of Hg were equivalent in magnitude to losses of dry and C mass, pointing to potentially strong release of Hg upon litter decomposition. The observed correlations of Hg with the distribution of C in U.S. forest, along with observed evasion losses of Hg upon C mineralization, indicates important consequences of global change and predicated changes in terrestrial carbon pools on the storage of Hg in terrestrial ecosystems.

TS8-P3 — 11:00-12:00 and 17:30-18:30
Authors: PORVARI, Petri1, VERTA, Matti1, LAURÉN, Ari2, LINJAMA, Jarmo1
(1)Finnish Environment Institute (SYKE), petri.porvari@ymparisto.fi; (2) Finnish Forest Research Institute;

We studied the effects of forestry practices on total mercury (TotHg) and methyl mercury (MeHg) concentrations in outflow stream and on output fluxes from a small (0.071 km2) boreal spruce forest catchment in southern Finland. (61°01’N, 24°45’E) using paired-catchment experiment. The soil in the treatment catchment was composed of coarse sand till overlain by a thin humus layer. Approximately 5 % of the treatment catchment area was covered by peat soil. The dominant tree species forest stand was Norway spruce. The treatment catchment was clear cut in 1997, mounded in 1998 and afforested in 1999. Runoff was measured continuously from 1994 to 2009 and discrete water samples for analysis of TOC, TotHg, MeHg, SO4 and general stream water chemistry were collected from weirs at the outlet of the treatment and the reference catchments.

We observed significant increases in the outflow of TotHg and MeHg from a forest catchment after clear cutting. The annual runoff ranged only between 50-280 mm in the control catchment, while the rainfall in the region was 450-710 mm indicating high filtration to deep soil layers and low surface water outflow from the untreated catchments. Clear cutting increased the runoff from 14 to 180 mm. The background total Hg export ranged between 0.25 and 1.70 g km-2 a-1, while the export of MeHg was 0.02-0.11 g km-2 a-1. The TotHg export increased from 0.25 to 4.2 g km-2 a-1 and MeHg export from 0.07 to 0.50 g km-2 a-1 in the years after treatment. The MeHg increase was significant in all years but TotHg only in the first year after the treatment. Further, the highest surplus of MeHg output was measured during the five last years of monitoring and indicated no signs of recovery to background levels. We will discuss the fundamental and long lasting effects on environmental characteristics that may be developed after clear cutting and soil treatment and lead to MeHg formation in forest soils.

There was no clear leveling off of MeHg concentrations after twelve years of the treatment, which necessitate further monitoring. In recent years there has been extensive felling of timber in Finland. This may cause elevated fish Hg levels in wide areas with forest lakes.

TS8-P4 — 11:00-12:00 and 17:30-18:30
Authors: EKLÖF, Karin1, KRAUS, Andrea1, MEILI, Marcus2, BISHOP, Kevin1
(1) Department of Aquatic Sciences and Assessment, karin.eklof@slu.se; (2) Stockholm University;

Forestry operations have been reported to increase the levels of both total mercury (THg) and methylmercury (MeHg) in runoff water and downstream biota. However, the sensitivity to forestry among sites seems to vary; while some studies have resulted in dramatic effects on THg and MeHg concentrations, other studies resulted in no or only moderate effects. A synoptic approach, covering a wide range of field sites in a spatial scale subjected to different forestry operations, could potentially contribute to understanding what determines this variation in sensitivity towards forestry. In this study we will investigate the individual influence of stump harvest and site preparation for replanting on the concentrations of total mercury (THg) and methylmercury (MeHg). We also evaluated the influence of treated areas (subjected to either stump harvest or site preparation) compared to untreated reference areas. Stump harvest is a forestry method used to increase the supply of bio energy from forests. However, it is important to compare the environmental benefits of increased biofuel supply with the possible ecological risks that the disturbance may cause. Stream water draining 55 catchments subjected to stump harvest (n=16), site preparation (n=21) or untreated forest (n=18) were sampled on two occasions and analyzed for THg, MeHg, total organic carbon (TOC), absorbance (Abs420) and particle concentration. The sample sites were spread all over Sweden. Preliminary results from the first sampling occasion indicate that there were no differences of neither THg nor MeHg concentrations between the stump harvested and the site prepared sites. Although not significant, the mean concentrations of both THg and MeHg were higher in the treated areas (9.52±6.69 ng/L, 1.51±1.90 ng/L) than in the reference areas (6.43±3.87 ng/L, 0.79±0.86 ng/L). The high standard deviation of the THg and MeHg concentrations in the treated areas indicate that site specific catchment characteristics play an important role in the overall THg and MeHg mobilization. The poster will present the evaluation of the full data set from the two sample occasions as well as an evaluation of site specific factors affecting the concentrations of THg and MeHg.

TS8-P5 — 11:00-12:00 and 17:30-18:30
Authors: ROSS, Donald S.1, JUILLERAT, Juliette1
(1) University of Vermont, dross@uvm.edu

We evaluated the importance of forest type on total Hg (THg) flux in litterfall and on THg accumulation in organic soil horizons. Eighteen actively managed forest sites were sampled throughout Vermont, USA. Mercury concentration was measured in senescing leaves of dominant tree species (16 species in total) in three forest types (low-elevation coniferous, deciduous/northern hardwood forest and mixed). Upper soil horizons (forest floor and A horizon if present) were sampled and analyzed for both THg and carbon. Mercury concentration in senescing leaves varied significantly between tree species. Some hardwood species had higher THg concentration in leaves than coniferous species. Acer pensylvanicum (striped maple) had significantly higher THg concentration than any other species. Total Hg concentration was negatively correlated with leaf height on the tree. Leaf surface-to-weight ratio was positively correlated with THg. The calculated THg flux in litterfall varied from 12 to 28 ug/m2/yr and there were no significant differences among forest types. Results showed an unexpectedly high THg concentration in understory trees. Similar to other studies, soil THg concentration was lower (73 ng/g) in the Oi horizon (litter layer) and peaked (226 ng/g) in the Oe horizon (fermentation layer) before declining (181 ng/g) in the humified Oa and/or A horizons. We compared THg pools to a depth of 9 cm from the surface litter layer, regardless of the soil horizons present (9 cm was the deepest common sampling depth for all sites). Mineral soil A horizons tended to have greater THg pools, up to 185 g/ha, than the forest floor due to their greater bulk density. Because of this, sites with thinner forest floors had greater 9-cm THg pools. Factors that controlled forest floor depth, e.g. thicker with higher elevation or thinner with steeper slopes, also related to this near-surface THg pool. Forest floors were especially thin where there was legacy from agricultural land-use resulting in the presence of both an Ap horizon and earthworms. Near-surface (9-cm) THg pools were highest on these sites. Interestingly, we did not find a relationship between the amount of THg in litterfall and that accumulated in the near-surface soil. Our overall results suggest that the assumption that needles have greater THg concentration than leaves may need to be revisited, with forest structure taken into account.

TS8-P6 — 11:00-12:00 and 17:30-18:30
Authors: BÉLIVEAU, Annie1, LUCOTTE, Marc1, DAVIDSON, Robert2, PAQUET, Serge1
(1) Université du Québec à Montréal, beliveau.annie@gmail.com; (2) Université du Québec à Montréal, Biodôme de Montréal;

Previous research has shown that the conversion of the Amazon forest into agricultural lands through slash-and-burn practices, causing the release of natural soil mercury (Hg) towards aquatic environment, has important impacts on ecosystems and human health. Local communities are exposed to this contaminant via fish consumption. As recent studies suggested that soil Hg mobility could be related to the massive cations input resulting from biomass burning, the adoption of agroforestry systems based on a non-recurrent use of fire could be an alternative to traditional slash-and-burn cultivation. Our objective was to compare soil Hg changes and distribution one year after forest clearing, in two types of cultivation systems, as an effort to encourage the adoption of more sustainable land-use practices in the Amazon. Our results showed that, at initial state (before cultivation), no regional gradient existed in soil Hg levels and that clayey soils had higher total Hg levels (most Hg was adsorbed on soil fine particles). An inverse relationship between initial cations and Hg was also observed, suggesting a competition for adsorption sites in soil fine particles. One year after forest clearing, in both types of agricultural systems, signs of Hg mobility was detected at soil surface (0-5cm). In slash-and-burn cultivation sites, a shift in Hg soil fractions occurred at surface horizon. Indeed, a transfer of Hg from fine to coarser soil (and less reactive) particles was observed, indicating that chemical bonds between Hg and fine particles could have been altered. This early Hg mobility in the soil suggests that Hg release would probably occur as agricultural activities go on. In agroforestry systems, Hg loss was observed in surface horizon for clayey soils and in deep horizon for sandy soils. This Hg loss occurred in first year of cultivation, reflecting similar patterns of those observed in traditional agriculture at this early stage, which could be explained by the use of fire to implement the agroforestry plantations. This study reinforces the assumption that the act of burning is of primary importance for soil Hg release, and justifies promoting agricultural practices based on a reduced use of fire in order to lower negative impacts of land use on soils, ecosystems and human health.

TS8-P7 — 11:00-12:00 and 17:30-18:30
Authors: ROGNERUD, Sigurd1, FJELD, Eirik1, KJÆR, Roar2
(1) NIVA, sro@niva.no; (2) Hedmark County Govenor.

Mercury contamination of freshwater fish is a severe environmental problem in Norway, and the health advisory maximum level of 0.5 mg/kg is often exceeded in boreal lakes. The deposition of long-range transported atmospheric mercury pollution is the main source. Most of the deposited mercury is stored in the soils and mobilization to surface waters can be a significant source for mercury in fish. Forestry and manipulations of forest soils are suggested to increase the leakage and elevate mercury levels in aquatic biota.

To investigate if the areas of clear-cutting and soil disturbance in lake catchments are connected to elevated mercury levels in perch.

Material and Methods:
Mercury concentrations, together with stable N- and C- isotopes and primary individual data, were recorded for perch (N=370) from 18 boreal lakes, sampled in 2009. The clear-cutting areas in the catchments were mapped and their ages determined by air photos and GIS based tools. The influence of clear-cutting, together with other environmental characteristics, were investigated by different statistical tools.

Results: No statistical significant effect of clear-cutting could not be detected by a mixed models approach. However, in a multiple regression model the mercury concentration in 3-years old fish were dependent on the percentage of clear-cutting in the catchments, lake surface to catchment ratio, and sulfate concentration in lake water. By a redundancy analysis we explored the multivariate relationship between mercury in fish and environmental variables.

We found a statistical significant effect of clear cutting on 3-year old perch, together with lake surface to catchments ratio, and sulfate concentration. The influence of forestry is probably masked in older fish by the accumulating effect of confounding variables. The negative effect of sulfate on fish mercury is surprising, and we discuss possible mechanisms in light of recent findings on how sulfur chemistry controls transport of mercury forms and methylation.

TS8-P8 — 11:00-12:00 and 17:30-18:30
Authors: OESTREICHER, Jordan1, LUCOTTE, Marc1, ROMANA, Christina2, ROZON, Christine1, MOINGT, Matthieu1, BÉLANGER, Emilie1
(1) GEOTOP-UQÀM, jsoestreicher@gmail.com; (2) ;

In the Tapajos River Region of the Brazilian Legal Amazon, land-management practices and extensive deforestation have been identified as primary factors leading to the transport of naturally formed soil mercury from the pedosphere to the hydrosphere. In this same region, river-dwelling populations are exposed to elevated levels of mercury through frequent consumption of fish that bioaccumulate the toxic organometallic compound. Exploring the causal relationships between punctual and prolonged land-cover changes and ecosystem responses can inform our understanding of mercury transport processes and mercurial exposure dynamics in riparian populations. Given this context, the present study applies a coupled analysis of spatio-temporal landscape changes and vertical mercury deposition patterns. Landscape metrics in seven lake catchments in the Tapajos region were determined through of spectral analysis and photo interpretation of three Landsat TM images over a twenty-year period (1987 – 2009). A geochronology of lacustrine mercury deposition records, using lead-210 activity analyses, was determined for sediment cores (70-100cm) taken at the outlet of each catchment. Preliminary comparisons of catchment-level landscape metrics with long cores suggest that the velocity of land cover change and the spatial extent of forest, pasture, and agriculture covers affect mercurial deposition patterns and sedimentary mercury burdens.

TS8-P9 — 11:00-12:00 and 17:30-18:30
Authors: CRANE, Sharron1, HUSAR, Michael1, DIGHTON, John1, BARKAY, Tamar1
(1)Rutgers University, shahicks@rci.rutgers.edu

In wooded ecosystems like the New Jersey (NJ) Pine Barrens where the buffering capacity of soil is low, atmospherically deposited mercury (Hg) is highly bioavailable. Also, Hg deposited to the canopy may be adsorbed to leaf surfaces or taken up through stomates. Saprotrophic fungi are important members of the soil microbial community and contribute to soil processes such as bulk litter decomposition and nutrient mobilization. Because of their potential role in both ecosystem response to Hg and mobilization of Hg in terrestrial ecosystems, a better understanding of the responses of soil saprotrophic fungi to Hg is needed. Here we report the initial results of experiments designed to elucidate the effects of Hg on soil saprotrophic fungi. Five soil samples from the NJ Pine Barrens were subjected to particle filtration and subsequent spread plating on MEA agar with added Hg concentrations ranging from 0-175 µM. Over a five-day incubation period, new colonies were enumerated and characterized according to morphotype. Few colonies grew on plates with added Hg concentrations higher than 50 µM. The diversity of fungi was significantly lower on MEA+50 µM Hg plates than on MEA plates: principal component analysis suggested that different members of the fungal community were dominant on each of these two agar types, this was supported by inspection of the raw data. Twenty-one unique morphotypes were found, and a representative colony of each morphotype was chosen for isolation. Four isolates were investigated for Hg tolerance: of these, two were tentatively identified as Umbelopsis isabellina, one as U. rammaniana, and the fourth as Aspergillus cervinus. The effect of Hg on the radial growth of the three Umbelopsis spp. was slight and not significant: all three were tolerant of Hg up to 20 µM. However, radial growth of A. cervinus was 20% inhibited by 20 µM Hg: this inhibition increased to 50% at 35 µM Hg. These results suggest that varying degrees of fungal tolerance to Hg may lead to changes in fungal community composition in forest soils that are impacted by Hg. Because litter decomposition plays a critical role in controlling Hg transport, the varied Hg tolerance of soil saprotrophic fungi may affect Hg geochemistry in wooded ecosystems.

TS8-P10 — 11:00-12:00 and 17:30-18:30
Authors: BHATTI, Jagtar S. 1, NASR, Mina2, ARP, Paul A. 2, NOSEWORTHY, Josh2
(1) Canadian Forest Service Natural Resources Canada, Jagtar.Bhatti@NRCan-RNCan.gc.ca; (2) University of New Brunswick;

Forests are found as significant scavengers and storage of mercury (Hg) in watersheds. This study aimed forest biomass modelling to quantify the extent of accumulation and release of mercury in softwood and softwood-mixedwood stands. Within this modeling framework, atmospheric Hg scavenged by the forest canopy is allocated to the foliage, wood, forest floor and roots components. Also considered is the turn-over of the accumulating forest biomass and biomass-attached Hg into soil organic matter, and the recycling of Hg that is already present in the soil, assuming steady-state conditions. Soil weathering rate, litterfall, decomposition rate, forest floor and soil biomixing, fire cycle, and surface evasions including surface volatilization and fungi and ground vegetation uptake are included in the model.

TS8-P11 — 11:00-12:00 and 17:30-18:30
Authors: NASR, Mina1, BHATTI, Jagtar S.2, ARP, Paul A. 1
(1) University of New Brunswick, mnasr@unb.ca; (2) Canadian Forest Service Natural Resources Canada;

In this study, litter (L-), fermentation (F-), and humification (H-) layers of forest floor, eluviation (A-) layer of top mineral soil samples were conducted from selected forests on Grand Manan Island, the south-west shore region at Lepreau and New River Beach, and the interior at Fredericton, of the province of New Brunswick, Canada. This study revealed that total mercury (THg) concentration in the organic soil was strongly correlated with the sulphur (S) content of the organic soil, and in the mineral soil it decreased with decreasing soil carbon (C) content. The THg/C and THg/N ratios of the L-, F-, H-, and A-layers appear to be functionally related to the state of organic matter humification, the slower turnover rates of S and N versus C, and of Hg versus S and N. The generated multiple regression analysis revealed that THg concentrations in the soil were the highest in F- and H- layers and it increased in area partially or completely covered with Pleurozium schreberi and Sphagnum sp. The model also showed negative correlation between THg concentrations and substrate thickness (soil depth and moss height), and C content, however positive correlation were observed for S content, substrate thickens, location, and soil layer type (F- and H- layers).

TS8-P12 — 11:00-12:00 and 17:30-18:30
Authors: FOSTIER, Anne Hélène 1, FORTI, Maria Cristina2, JARDIM, Wilson F.1, CARVALHO, João 3
(1) UNICAMP, fostier@iqm.unicamp.br; (2) INPE; (3) UNESP/INPE.

In the Amazonia, as in other forest ecosystems, mercury is present as a result of natural mineral deposits, atmospheric deposition, and regional anthropogenic sources of pollution, including gold mining. Intact forests and soils within these systems serve as a sink for mercury, diminishing or delaying its movement in the environment and in the human food chain. In the Amazonian region deforestation rate was around 1.7 x 106 ha yr-1 between 2000 and 2010 and biomass burning is generally used as the first step of deforestation. How this high rate of deforestation and biomass burning can impact mercury cycle was the aim of several studies conducted in different places of the Amazonian region (Negro river basin and different areas of the states of Amapá, Mato Grosso and Acre) by our laboratory over a period of 10 years. Mercury concentrations were determined in many compartments (air, rainwater, throughfall, river water, sediment, fresh vegetation, litterfall, soil, fish and human) allowing to calculate and to compare Hg fluxes at the atmosphere/vegetation, vegetation/soil, soil/atmosphere (deposition and emission) and soil/water interface in forests and deforested areas. The factors involved in the mercury immobilization/mobilization processes observed in forests and deforested areas were also studied. This paper aims to present a synthesis of these studies.

Tuesday, 26 July, 2011